Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Bretton Woods system wikipedia , lookup
International status and usage of the euro wikipedia , lookup
Reserve currency wikipedia , lookup
Foreign exchange market wikipedia , lookup
Currency War of 2009–11 wikipedia , lookup
Foreign-exchange reserves wikipedia , lookup
Currency war wikipedia , lookup
Purchasing power parity wikipedia , lookup
Fixed exchange-rate system wikipedia , lookup
International monetary systems wikipedia , lookup
Chapter 6 ACCEDING COUNTRIES: THE ROAD TO THE EURO 1. Introduction The ten acceding countries are expected to join the euro area at some point in the next few years. Their participation in the monetary union is seen as the end point of their process of integration in the EU. In the “pre-accession phase”, the acceding countries have fulfilled the acquis communautaire in the area of EMU: they have implemented reforms to make their institutions compatible with joining the European Union (for example, they have made their central banks independent) and liberalised capital flows. At the start of the “accession phase”, they are expected to pursue policies aimed at fulfilling the nominal convergence criteria established by the Maastricht Treaty and qualify for joining the euro area. The third and final phase will start with the formal replacement of domestic currencies with the European currency. The EU treaties do not contemplate the possibility that acceding countries opt out of the euro: joining the EU automatically requires these countries to take the necessary steps to enter the EMU. The Maastricht Treaty lays out the conditions and procedures to become a member of the euro area, which are the same for both old and new members of the EU. Technically, acceding countries will join the European Union with “a derogation”. Thus, the only policy decision by acceding countries is whether to try to join EMU at an early or a late stage after accession. In principle, a country could choose to delay participation in EMU indefinitely. In this respect, it is useful to recall that, contrary to the United Kingdom and Denmark, Sweden is staying outside EMU without having negotiated any “opt-out clause”. But the perspective of postponing EMU participation does not seem to be appealing to current governments in acceding countries: all of them have declared their willingness to adopt the euro as soon as possible. The traditional argument in favour of EMU participation is credibility of low inflation, which applies to the newly acceding countries as it did to the Southern European countries in the 1990s. A common currency eliminates currency risk and reduces interest rate differentials. Such credibility gains are an advantage from Box 6.1 The nominal convergence criteria • • • • Deficit of the general government must be below three percent of GDP. Gross debt of the general government must be below 60 percent of GDP, or declining toward 60 percent of GDP at a satisfactory rate. Inflation rate must not exceed the average rate in the three EU countries with the lowest inflation rate by more than 1.5 percentage points. Long-term interest rates must not exceed the average rate in the three EU countries with the lowest inflation rate by more than two percentage points. Two years of participation in the Exchange Rate Mechanism II (ERM II) without major tensions in the foreign exchange market. ERM II replaced the ERM of the European Monetary System created in 1979. ERM II was established in 1997 with a resolution of the European Council in order to link the currencies of EU member states outside the euro area and the euro. Like ERM I, ERM II is also a multilateral exchange rate arrangement with a fixed, but adjustable, central parity and a fluctuation band around it. Countries participating in ERM II peg their exchange rates to the euro, allowing for fluctuations within a symmetric band of 15 percent on each side of the central parity. Interventions at the margin are automatic, unless they conflict with the primary objective of price stability in the euro area. Decisions concerning central parity adjustment, or fluctuation within margins narrower than 15 percent, are taken by mutual agreement. Any member (including the ECB) can initiate a confidential procedure to reconsider central rates. As established by the Ecofin Council in 2000, currency boards are compatible with ERM II participation. The following regimes are incompatible with ERM II: floating, crawling peg, peg against a currency different from the euro, and unilateral euroisation (Ecofin Council 8/11/2000). Participation in ERM II is voluntary, but acceding countries are expected to join as a precondition to adopting the euro. The procedure to join ERM II can be initiated at any time by a EU member state. There are no formal criteria for joining ERM II. Whether a country satisfies the convergence criterion of exchange rate stability will be judged by looking at a range of indicators. As discussed below in the text, a country is expected to keep its exchange rate close to the central parity in ERM II or experience a currency appreciation. The assessment will also take into account short-term interest rate differentials and the frequency and intensity of foreign exchange market interventions. 119 EEAG Report Chapter 6 the point of view of macroeconomic stabilisation that should be set against the loss of national monetary policy as a stabilisation instrument and of exchange rate flexibility as an adjustment mechanism.1 Recent literature has stressed a number of effects of joining EMU which are not analysed in the traditional debate. First, a common currency is likely to increase trade within the EU. In this respect, adopting the euro is equivalent to a drop in transaction costs in cross-border exchanges of goods and services within the EU economic area. Second, by reducing the stock of external debt denominated in a foreign currency, adopting the euro will substantially reduce vulnerability to currency and financial instability (although in principle EMU countries could still issue large stocks of dollar-denominated debt). We will discuss this point at length below. Taking it as a political fact that the accession countries will ultimately join EMU, the question of the optimal timing of such a move becomes crucial. The timing directly impinges on the acceding countries’ ability to stabilise their economies in the next few years and build an economic environment that favours high rates of investment and growth, economic integration and financial stability. Fiscal and monetary authorities in acceding countries now operate in a regime of high capital mobility. This is the result of a relatively rapid process of liberalisation and deregulation implemented in the last few years. But their domestic institutions and markets have only recently started to operate in a fully liberalised and deregulated system. Whether or not the financial and legal systems of these countries can weather volatile capital movements is perhaps too early to say, but it would be naïve to hope for the 1 The traditional theory of optimal currency areas (OCA) suggests a set of criteria to judge the costs of giving up exchange rate flexibility. These include the degree of price and wage flexibility, the extent of factor (labour) mobility, insurance via fiscal transfers or fiscal integration in general, and correlation of macroeconomic non-financial shocks. This last item can be attributed to several structural features of the macro economy: openness and economic size, degree of goods market integration, composition of production and trade specialization. Note that, per se, correlation of business cycles is not an OCA criterion, as a high correlation may not result from symmetric shocks but from symmetric policy responses to asymmetric shocks. Actually, national business cycles in Europe may well become more correlated after the creation of EMU, without implying that the cost of giving up exchange rate flexibility has fallen at all (see Corsetti and Pesenti 2002). The literature applying these criteria to judge whether acceding countries are ready for EMU is extremely vast, but its results are quite inconclusive (at least as inconclusive as the results of the literature on the same subject applied to EMU creation). The specific problem with the OCA approach applied to acceding countries is how to account for structural changes currently under way in the convergence process. We believe that these changes are exceedingly difficult to assess, and therefore provide shaky ground for empirical exercises. EEAG Report better and envision years without large (global or region-specific) shocks. As discussed in Chapter 5, structural imbalances in these economies may cause acute problems. Deteriorating fiscal conditions could constrain the use of budget policies for stabilisation purposes. Stabilisation is likely to fall disproportionately on monetary and fiscal authorities, both from a macro perspective and from a financial stability perspective. In such an environment, mandatory adoption of a regime of limited exchange rate flexibility (the ERM) for two years before entering EMU is quite controversial and has stirred a considerable debate in both policy and academic circles (see Buiter and Grafe 2002, Begg et al. 2003 and ECB 2003 among others). In this chapter, we will reconsider the main issues related to the choice among alternative paths of transition to the euro. Independently of the exchange rate regime, a high degree of capital mobility in the transition to the euro will entail high financial risk. The choice of inappropriate exchange rate regimes can, however, magnify this risk greatly. The experiences from recent crises and financial turmoil in emerging markets show that wrong decisions can be extremely painful – the punishment for apparently small mistakes can be enormous. Based on this experience, we will try to single out the policy that could reduce vulnerability to currency and financial instability. We will devote a large part of our discussion to the debate on the costs and benefits of participating in “ERM II”, which is the Exchange Rate Mechanism with large fluctuation bands. However, the exchange rate regime is only one dimension of the policy framework. What is ultimately important for acceding countries is pursuing the right stabilisation policies, strengthening their fiscal, financial and monetary institutions, making their economies flexible and choosing a consistent level of social protection. For the EU, policy priority should be given to strengthening its financial architecture, along the lines extensively discussed in Chapter 4 of the 2003 EEAG report. 2. Exchange rate regimes with liberalised financial markets: the current policy framework of acceding countries In Chapter 5 of this report, we have seen that the acceding countries are rapidly integrating their markets for goods and services in the world and the EU 120 Chapter 6 economies. Acceding countries are highly open – the average degree of openness (exports plus imports as a share of GDP) is around 100 percent of GDP, compared with 70 percent of the present EU countries (Poland and Cyprus are significantly below average, however). Approximately 60 percent of the acceding countries’ imports and exports is with EU countries. A significant share of EU trade with the acceding countries is intra-industry, clearly reflecting cumulated foreign direct investment (FDI) by EU countries. Openness is destined to increase after accession. Thus, these countries will be highly exposed to terms of trade shocks and international demand spillovers from both within and outside the EU area. Acceding countries have removed most restrictions on capital mobility, with the exception of restrictions on the purchase of land by foreigners. Slovenia also keeps some controls on short-term capital and direct investment. Financial integration with the EU is already quite high. Over the last few years, two thirds of capital inflows to acceding countries consisted of FDI flows: approximately 80 percent of these flows originated in the EU. Some indicators show that the financial systems of the acceding countries are moving towards the EU performance standard (see European Commission 2003). In Cyprus and Malta the size of the financial sector is close to the EU average. But compared to the rest of the EU, the domestic financial sector in the acceding countries in Eastern Europe is still underdeveloped. Banks dominate the financial sector. In the last few years, most banks were privatised – only in Poland and Slovenia do the governments still retain ownership of some large financial institutions. The privatisation process has coincided with a change in ownership from domestic to foreign. Figure 6.1 in the initial stages of their life (European Commission 2003). The room for financial deepening in the Eastern European acceding countries is large. In the next few years, credit to both households and corporations, stock market capitalisation as well as activities by financial intermediaries will probably experience very rapid growth. Financial deepening can be extremely beneficial: it can relax credit constraints, provide a sufficiently diversified supply of funds to finance projects with different risk profiles, and create opportunities for risk diversification and the reallocation of consumption over time. Yet, a high speed of expansion can also cause a deterioration of allocative efficiency if it leads to excessive risk-taking, potentially undermining the contribution to welfare of financial market development. Country size, capital mobility and exchange rate regimes Figure 6.1 plots the exchange rates of the acceding countries, while Table 6.1 reports the exchange rate regime adopted in 2003 as well as changes that have occurred since the beginning of the 1990s. Exchange rate regimes are classified according to the official IMF classification allowing for some suggestions by various researchers (see von Hagen and Zhou 2002 among others). There are two striking features in Table 6.1, already noted by many commentators. The first is a positive correlation between country size and the flexibility of the exchange rate regime. Looking at 2003, smaller countries like Estonia and Lithuania have currency boards (arrangements that constrain monetary In 2000, domestic credit amounted to 60 percent of GDP in the acceding countries, against an average of 140 percent in the euro area. Gross debt of individuals was quite low: the average for the acceding countries was seven percent of disposable income, against an average of 50 percent for the euro area. Stock market capitalisation was also low. Several indicators show lack of funding for small and medium-sized firms, especially 121 EEAG Report Chapter 6 Table 6.1 The evolution of exchange rate regimes in acceding countries 1990 1995 2000 Cyprus 2003 Peg to the euro (±15% band, de facto ±1-2% band) Czech Republic Peg Peg Managed float Free float, inflation targeting Estonia Currency board Currency board Currency board with peg to the euro Latvia Peg Peg Peg to SDR (euro weight is 29%; ±1% band) Lituania Peg Currency board Currency board Currency board with peg to the euro Hungary Peg Crawling bands Crawling bands Peg to the euro (±15% band), inflation targeting Malta Peg to currency basket (±0.25% band) Poland Peg Crawling bands Crawling bands Free float, inflation targeting Slovakia Peg Peg Managed float Managed float Managed float Managed float Managed float Slovenia Note: IT stands for inflation targeting Source: Von Hagen and Zhou (2002); Begg et al. (2003). authorities to expand domestic money supply in line with international reserves), Latvia pegs its currency to a basket of international currencies within very narrow bands. Malta pegs to a basket of currencies with a 70 percent euro share, also within very narrow bands. Cyprus pegs to the euro, officially adopting large bands of fluctuation, but de facto pegging within extremely narrow bands. Conversely, larger countries adopt more flexible regimes. The Czech Republic and Poland have opted for a free float – central banks in these countries have adopted inflation targeting as their monetary strategy. The Slovak Republic and Slovenia pursue some form of managed exchange rate float. Hungary has adopted a regime with somewhat limited exchange rate flexibility combined with inflation targeting. Fixed exchange rate regimes provide a nominal anchor to pin down the price level in the economy. For the peg to be viable, domestic prices cannot move too far from international prices, because real exchange rate changes would have destabilising effects on aggregate demand. The choice of a more flexible exchange rate regime, instead, raises the issue of choosing a nominal anchor to guide private sector expectations of inflation. The largest acceding countries have opted for some form of inflation targeting – requiring monetary authorities to set explicit, yet contingent, goals in terms of rates of increase of the consumer price index (CPI). EEAG Report 122 The second feature of Table 6.1 is a tendency of governments to move away from intermediate regimes of limited exchange rate flexibility after capital market liberalisation. Indeed, most of the acceding countries had adopted some form of soft peg at the beginning of the 1990s. After liberalisation of capital flows during the 1990s, they have moved either to flexible exchange rate regimes, or to hard pegs/currency boards that in principle exclude any realignment. In 2003, five countries have a currency board or a hard peg, whereas five have a free float or some form of flexible regime. Free capital mobility and fixed exchange rates: inconsistent policy regimes? According to a view that has gained more and more supporters after the global turmoil of the 1990s, free capital mobility is inconsistent with any form of limited exchange rate flexibility. The reason is as follows. Rates of return from short positions in currencies that fall by 10, 20 or 30 percentage points in a short time-span are extremely high. With high capital mobility, taking a speculative position against a currency has low costs: investors will miss no opportunity to test the resolve of governments in maintaining their exchange rate target. As long as market participants think that the government can “realign” the exchange rate (that is, devalue the currency), the targeted parity will not be completely credible. Investors know that, depending on the circumstances, policy makers will prefer to modify the Chapter 6 exchange rate rather than to sustain the enormous costs of extreme defence. These costs possibly involve protracted periods of sky-high interest rates, with devastating effects on the health of the banking and financial system, and therefore on investment, growth and employment. Many factors determine the circumstances that may force governments to devalue. Some of them have a domestic origin, but very often the shock-triggering waves of speculative attacks originate in the global economy. Recurrent examples include monetary shocks in the OECD countries (higher US interest rates), as well as turmoil/crises in other countries or markets. What makes the problem worse is that speculative attacks may be driven by self-fulfilling beliefs. If investors consider an exchange rate policy credible, they will have no reason to speculate against the currency, and the current level of the exchange rate will be viable. Otherwise, they will launch speculative attacks that will undermine the stability of the peg. The ensuing currency and financial collapses validate ex post the initial expectations of turmoil. Different factors and events determine the ability of domestic policymakers to guarantee stability of their domestic markets and financial institutions, but the country’s vulnerability to them ultimately depends on its fundamentals. Sound domestic policies and strong institutions can eliminate the possibility of self-fulfilling crises. Low credibility of stabilisation policies and weak institutions cannot. 3. Vulnerability to crises: lessons from emerging markets As extensively documented in Chapter 5 of this report, the acceding countries are emerging markets: their income levels are considerably lower than in the rest of Europe, and their economies are growing rapidly. What lessons can we learn from the recent experience with macroeconomic stabilisation and exchange rate regimes of other emerging markets? The 1990s are rich in cautionary tales. During this decade, several factors contributed to the rapid growth of markets for emerging market assets, including low interest rates in the industrialised countries as well as the growth of financial instruments and the diversification of financial institutions. But, as is well known, severe financial turmoil hit 123 Mexico, Thailand, Indonesia, Korea, Russia, Brazil, Ecuador, Turkey, Argentina and Uruguay, to name the main ones. Before a crisis, these countries had all liberalised their capital accounts (although to different degrees) and adopted policies of low inflation and macroeconomic stabilisation. In addition, all of them had struggled to reduce their high country-risk premia charged in international capital markets. In all these episodes, a common pattern emerges. Capital mobility is associated with high volatility of capital flows. After capital account liberalisation, emerging markets pursuing macro stabilisation programs typically experience a very large upsurge of capital inflows. Such inflows are driven not only by high rates of return in economies with relative scarcity of capital, but also and to a large extent by the attractiveness of short-term profit opportunities from speculative positions. This is because, just after capital account liberalisation, country-risk and/or inflation tend to keep domestic interest rates high relative to international rates. Gains in the credibility of domestic policies drive the process of “convergence” of domestic to international interest rate levels over time, but usually at a slow pace. Large capital inflows into the country can just as easily and suddenly stop, opening up enormous financing gaps (see Calvo 2003). Convergence can create vulnerability When a country commits to a peg, its exchange rate, riding the wave of the “convergence process” is very attractive from the point of view of international fund managers and investors. These invest in shortterm debt issued by the country at high interest rates, expecting stable or appreciating exchange rates. When debt is denominated in foreign currency, the interest rate typically includes a substantial countrypremium over international rates. Speculative positions are invariably short-term, since investors prefer to be able to withdraw quickly from the country if there is any sign of trouble (i.e., rumours about devaluation or default. What we have described above is labelled “convergence play” in the literature on currency and financial crises. It has been common in most episodes of exchange rate-based stabilisation in emerging markets as well as in advanced economies. The “convergence play” became notorious in Europe during the period preceding the currency crises of 1992–93, when fund managers massively bought high interest- EEAG Report Chapter 6 rate assets denominated in, say, Italian lira, sometimes pretending to cover their positions by selling short assets denominated in D-marks (see Buiter, Corsetti and Pesenti 1998). tunities. Even if the external debt of acceding countries now reflects to a significant extent the cumulated stock of FDI, the composition of external debt may change rapidly in the next few years.2 In emerging markets, large inflows driven by the “convergence play” are problematic in many dimensions. First, by feeding spending on both consumption and investment goods, they usually lead to an overheating of the economy. Second, by expanding the demand for short-term assets denominated in domestic currency at a rate several times higher than the growth rate of the economy, they provide a strong incentive for financial intermediaries to expand their activities without paying sufficient attention to prudential standards. To the extent that newly available funds drive up housing and land prices, the rising value of collateral assets further magnifies the incentive to create credit. Excessive credit creation exacerbates the fragility of the financial system (see Corsetti, Pesenti and Roubini 1999). In addition, with abundant capital inflows, policymakers perceive a softening of credit constraints. Not only does this create an incentive to borrow too much: abundant liquidity may also make governments more willing to extend public guarantees on private projects. Liberalisation and privatisation magnify the distortions due to public guarantees and “connected lending” (where credit is obtained through political links). This, too, contributes to excessive risk taking. Finally, and most importantly, since most debt is short-term and denominated in a foreign currency, the country is highly illiquid and exposed to destructive debt runs. Is foreign direct investment a solution? Are acceding countries likely to experience a “convergence play”? One may argue that these countries have already experienced large inflows of capital, most of which in the form of foreign direct investment (FDI). Moreover, interest rates in acceding countries are not far above the international rate. Why should the composition of capital flows change in the future? The problem is that participating in the EU will change the international assessment of risk in these countries. The EU “seal of approval” and the macroeconomic stabilisation programmes that accompany accession are likely to induce a new wave of capital inflows, this time with a stronger portfolio component. In other words, stabilisation and convergence policies are very likely to attract portfolio managers chasing short-term and medium-term profit oppor- EEAG Report 124 Suppose, however, that capital will still flow into these countries mostly in the form of FDI. Would this mean that these countries are sheltered from crises and/or their adverse consequences? Indeed, FDI flows have two major advantages over foreign debt as regards financial and currency stability. First, they are driven by real investment opportunities and therefore tend to be long-term. The data show that they are much more stable than portfolio flows. Second, the return on FDI depends on the profitability of real investment and is therefore pro-cyclical and ultimately contingent on the performance of the economy. In the presence of currency and financial turmoil, FDI investors typically suffer capital losses: the international value of the country’s liabilities drops in a crisis. Consider instead external debt denominated in foreign currency. Payments on debt are not contingent. In a crisis associated with a drop in output and devaluation, the burden of foreign debt increases: the larger the rate of devaluation, the sharper the revaluation of the country’s external liabilities. These two advantages of FDI over foreign debt clearly reduce the vulnerability to financial turmoil of countries with a large share of FDI in total capital inflows. However, the argument in favour of FDI is sometimes taken one step further. It is argued that, since FDI investors lose from a currency collapse, they will be unwilling to speculate against the country in which they have relatively illiquid assets (or perhaps that they will even be willing to take long positions in domestic currency in episodes of turmoil!). If this argument were true, large FDI investments could in principle shield a country almost completely against currency crises. Unfortunately, this argument is wrong. The point is that, once rumours of devaluation spread, investors who own domestic capital have the strongest incentive to hedge against capital losses due to the drop in the international value of their assets in the country. They will therefore take a short position against the 2 The recent default crisis in Argentina affected investors’ attitude towards lending to sovereign states and private firms in emerging markets. To the extent that investors will be reluctant to engage in “convergence play”, the crisis in Argentina may turn out to have some beneficial implications for acceding countries. Chapter 6 currency and/or the stock market index in the country. So, while FDI flows have indeed many properties that strengthen financial stability, FDI investors can also cause massive volatility in short-term capital flows as a result of hedging strategies by firms with capital in the country at the onset of a crisis! To put it simply, it is largely a myth that FDI can eliminate speculation in the currency and asset markets (see Guimaraes and Morris 2003).3 Currency and maturity mismatch When most flows from abroad are short-term and are directed towards assets denominated in international currencies, the maturity structure and currency denomination of the country’s external debt create a strong imbalance for the financial system as a whole. Unless firms, households and banks hedge their positions (the evidence is that they hardly do so), fluctuations in exchange rates and asset prices have strong effects on the balance sheets of domestic agents and institutions. Currency crises may turn into widespread bankruptcy in both the banking and the real sector. These problems are quite compelling in acceding countries, since their financial systems are already operating to a large extent in a foreign currency. 3 Aside from the volatility aspect, there are also doubts that FDI is the most efficient way to channel capital to emerging economies, as with FDI the benefits from financial liberalisation accrue only to very few firms, mostly (but not exclusively) in the tradable sector. Conversely, bank flows – intermediated by the domestic banking system – are in practice the only source of external financing for firms in the non-tradable sector and/or small firms. Heavy reliance on stable FDI flows can easily lead to bottlenecks and strongly imbalanced growth. Moreover, in many cases FDI is mainly motivated by tax-saving schemes adopted by multinationals. Figure 6.2 125 Table 6.2 Outstanding euro-denominated bank-deposits As % of total deposits Slovenia 37.9 Estonia 19.2 Turkey 15.6 Bulgaria 15.3 Latvia 11.6 Hungary 11.2 Romania 9.3 Slovakia 8.5 Czech Republic 7.6 Malta 6.6 Poland 5.3 Cyprus 5 Lithuania 1.9 Source: “Review of the International Role of the Euro”, ECB December 2002. Thus, liabilities and financial assets are already to a large extent euroised, and the degree of euroisation may be expected to increase in the future. Table 6.2 and Figure 6.2 provide some evidence on the extent to which liabilities in selected acceding countries are denominated in a foreign currency. Could banks in acceding countries insure against these balance sheet effects by lending domestically in foreign currency? In this case there would be no currency mismatch in the balance sheet of the banking system, since both assets and liabilities would be denominated in foreign currency. The problem, however, is that many domestic firms borrowing in foreign currency obtain a large share of their revenues from sales in the domestic market. To the extent that domestic prices do not adjust one to one with the exchange rate, nominal devaluation would worsen these firms’ balance sheets: given the value of their cash flow in domestic currency, the local-currency value of their debt would increase with a drop in the exchange rate. A high rate of bankruptcies of borrowing firms would directly and indirectly affect the health of financial intermediaries. In other words, euroisation of both assets and liabilities of the banking sector in the acceding countries does not solve the problem: EEAG Report Chapter 6 exchange rate risk is simply translated into default risk for the banks. Does foreign ownership of banks shield against crisis? Stabilising an economy on a fast convergence track: challenges to monetary policy Another often heard argument is that financial stability in acceding countries is not a concern since the financial sector in these countries is dominated by banks, and a large number of banks are owned by EU-based financial institutions, responsible to their home country’s supervisors and regulators. Thus, EU standards in supervision and regulation apply to a large part of the financial sector in acceding countries. This point is well taken, but the argument misses an important element. Good supervision and regulation may constrain excessive risk taking but cannot eliminate balance sheet problems and/or liquidity runs. Bank headquarters abroad cannot be expected to intervene and provide the necessary funds to their branches in the country should these experience difficulties or are hit by liquidity runs. The argument that the role of central banks as lenders of last resort is less crucial when financial intermediaries are owned by foreign institutions (weak on logical grounds) has been definitely proved wrong by the recent experience of Argentina. To put it simply, foreign bank owners do not guarantee any liquidity provision or capital injection. Most crises fall into the grey area between liquidity and solvency, so that it is difficult to assess whether additional funds will save a specific bank or simply be lost in a bankruptcy. But even when crises are close to those of illiquidity, systemic nation-wide contagion creates a coordination problem among liquidity providers. Either all foreign financial intermediaries provide liquidity and the crisis is prevented or no individual intermediary has an incentive by itself to bring more funds into the country, as its own contribution will be insufficient to avoid systemic financial collapse and will therefore be lost in the speculative run. The pattern of strong capital inflows, driven by “convergence plays”, is present in the experience of many countries, whether or not their governments commit to peg the exchange rate, and independent of the specific features of the exchange rate regime. However, commitment to a peg may exacerbate the intensity and the consequences of “convergence EEAG Report plays” and strengthen the expectations of public sector involvement in bailing out both financial and non-financial private enterprises. 126 In a boom with large capital inflows, monetary authorities have a hard time stabilising the economy. If they raise interest rates to reduce domestic overheating, they may exacerbate the “convergence play”. If they try to sterilise the capital inflow, reserve accumulation can become excessive by any standard. Nominal and real appreciation of the exchange rate, while reducing overheating, can harm exports and create the premise for currency and financial crises in the future. Regulatory institutions and supervisors have a hard time enforcing prudential standards, as price signals (for example, a high value of collateral) may create a misperception of risk. What is particularly disturbing for policymakers is that any policy aimed at slowing down an overheating economy may not be easily communicated to the public, since accession to the EU has created very optimistic expectations of fast convergence. How can one distinguish between an overheating economy and an economy that is simply on a fast convergence track? Efforts of policymakers to avoid overheating could be interpreted as misguided policy. In Chapter 5, we have shown that convergence of Eastern Europe will realistically take about 20 to 25 years in the best-case scenario for the most advanced countries. More likely it will take 50 to 100 years in most countries. What is not clear is whether cyclical stabilisation can be well defined and understood by the public in a country that is far away from its steady state growth rate. 4. Reducing vulnerability Recent experience with stabilisation and crises in emerging markets (including Eastern European countries) suggests two important considerations in designing stabilisation policies in acceding countries. Financial frictions and the size of the business cycle First, business cycles in acceding countries are likely to be more pronounced than in the EU and perhaps comparable to boom-bust-cycle episodes recently Chapter 6 experienced in emerging markets. An important reason for cyclical variability is the currency denomination of external debt. To the extent that debt is denominated in foreign currency, in periods of real exchange rate appreciation the value of liabilities in the banking system will fall, allowing banks to lend more. In periods of sudden reversal and depreciation, banks will be experiencing stress and/or crisis. Both cyclical upswings and downturns are likely to be more pronounced than in the current EU member countries (see Tornell and Westermann 2002). Early adoption of the euro can in principle eliminate this magnification mechanism of the business cycle amplitude. Yet, even after euroisation, there could be other financial market imperfections that would still contribute to large fluctuations in economic activity. Crises are not necessarily bad news for long-run growth Second, while convergence by the acceding countries is likely to be “bumpy” (that is, characterised by occasional episodes of financial turmoil and crisis), this is not necessarily bad for long-run growth. The main issue is whether, on average, a country could be better off by avoiding instability altogether, even if this happens at the cost of slowing down the process of deregulation/liberalisation of goods and financial markets and full economic integration in the European and world markets. Theory and evidence in this respect are not conclusive. For instance, on the one hand, risky balance sheets may lead to fragility and crisis; on the other hand, if economic agents are credit constrained, risky balance sheets are an (admittedly imperfect) way to overcome credit constraints during the boom phase of the cycle. Looking at the experience of emerging economies over the last 20 years, Tornell, Westermann and Martinez (2003) point out that, despite episodes of major crisis, Chile, Thailand and Korea are among the fastest growing economies in the developing world. In these cases, financial fragility notwithstanding, financial market liberalisation appears to be good for growth in the long run. This is not to say that crises are necessarily good for growth and that countries should disregard financial fragility in pursuing their growth process. But the experience of the last decade also makes it clear that there is no easy solution to the problem of reducing a country’s vulnerability to financial shocks. The following lessons apply. 127 A sustainable fiscal stance First, there is no monetary and exchange rate stability without a strong fiscal stance. One of the key factors underlying the crisis of Argentina’s currency board was the deterioration of fiscal conditions implied by a deep conflict between the central and the provincial governments. Such conflict undermined any expectation of fiscal discipline in the near and less near future. Even if current fiscal imbalances were still small at the time of the crisis, international investors and institutions came to realise that the path of Argentina’s public finances would not be consistent with exchange rate stability. In other words, current public deficits and the outstanding stock of public debt provide a partial picture of a country’s fiscal stance. Fiscal sustainability crucially depends on the market assessment of future and contingent liabilities. In many instances, this gives rise to the possibility of self-fulfilling crises. In anticipation of large deficits caused by a crisis, a speculative attack on a country’s assets can create macroeconomic imbalances that cause a crisis and thus validate ex-post the initial forecasts of fiscal troubles. Implicit and contingent fiscal liabilities are a crucial determinant of currency and financial fragility. These are magnified by inconsistent fiscal policy that feeds expectations of public bailout of firms in trouble and creates incentives for the private sector to take excessive risk. Well-functioning financial markets Second, there is no currency and financial stability without well-functioning financial markets. The main problem is that capital account liberalisation and deregulation magnifies the economic distortions associated with inconsistent financial policy. In practice, excessive risk taking means that, when undertaking projects, firms, households and banks believe that under some circumstances they will be able to avoid the bill if things go wrong – a bill involving both monetary and non-monetary costs. Reducing vulnerability does require strong financial regulation and supervision, which in turn involves the development of strong and efficient institutions. It also requires a process of privatisation of the economy accompanied by a clear definition of rules, laws and policies concerning risk management. In princi- EEAG Report Chapter 6 ple, this can be achieved in different ways, not necessarily through strict external regulation and supervision of financial firms but also through self-regulating bodies set up at industry level. But simply asserting the unwillingness by EU governments to bail out private financial and non-financial firms would not be effective, as such announcements lack credibility: investors know that the political pressure to intervene in crisis situations is very strong. In Chapter 4 of last year’s report we analysed the incentives for EU governments to intervene excessively at the national level. These incentives will become even stronger once the acceding countries are in the EU and even more so once they are in EMU. The tension between many national regulators and one monetary policy will be aggravated by the entry of new members with weaker institutional structures. An exit strategy from a fixed exchange rate regime Third, fixed exchange rate regimes suffer from an “escape clause” or “exit strategy” problem. Suppose a government adopts a hard peg as a means to reduce inflation. To be successful, the government needs to commit to keeping the current parity indefinitely. However, to the extent that the strategy is not immediately credible and/or there are contracts in the economy that predetermine inflation in the near future, the economy will suffer from increasing relative price imbalances during the first years of the strategy. High rates of inflation relative to the international rate will translate into a deterioration of competitiveness. By the same token, to the extent that low credibility translates into high risk premia in the asset markets, the stock of public debt will rise rapidly, whereas the capital stock will not (as investment will be relatively costly). At a given exchange rate, the correction of these imbalances requires the government to reduce domestic inflation below the international rate for some time, adding to the cost of disinflation and therefore raising the attractiveness of a nominal realignment of the exchange rate. But this possibility undermines the credibility of the policy strategy in the first place. Expectations of devaluation raise the costs of exchange-rate based stabilisation and can eventually become self-fulfilling. Either the exchange rate is kept fixed no matter what, or a fixed exchange rate policy becomes a recipe for crises. The longer the attempt lasts to resist devaluation, the greater is the magnitude of cumu- EEAG Report 128 lated economic imbalances. As the recent experience of Argentina shows, the bill of delaying the exit from unsustainable policies can become so large as to cause a collapse of prices, incomes and production. A high degree of wage and price flexibility Fourth, systems of inflexible exchange rates require some degree of price and wage flexibility. Those emerging market economies that weathered well the global shocks of the 1990s (such as Chile) experienced swings in the real exchange rate on the order of 20 to 30 percent over a business cycle. In these cases, nominal exchange rate flexibility has arguably provided the country with an extra degree of freedom to adjust to shocks. Countries pegging their currency also experience large swings in the real exchange rate, but without this extra degree of freedom to adjust to shocks. As the Argentinean government pegged the peso to the dollar, Argentinean exporters were exposed to large destabilising fluctuations in the peso value of the euro, as well as to fluctuations in the peso value of other Latin American currencies (mainly the Brazilian currency). The same can be said for the Asian countries that were pegging to the dollar or to a basket of currencies, in which the dollar had a substantial weight, at the onset of the crisis of 1997–98. The appreciation of the US currency in that case translated into a loss of competitiveness vis-à-vis Japan and other countries in Asia and Europe. The current dollar depreciation vis-à-vis the euro is likely to raise similar problems for acceding countries by inducing sizeable changes in their terms of trade. A consistent international financial architecture Fifth, vulnerability has a strong systemic dimension. The risk connected with maturity and currency mismatches in the external debt of emerging markets are understood by international investors, who are nonetheless willing to lend. Both domestic and international guarantees (in the form of international liquidity provision) induce creditors’ moral hazard: lenders have a weak incentive to differentiate among debtors of different quality if they believe that a combination of international bailouts and market timing (early withdrawal) can shield them against losses. The current debate on the reform of the international financial architecture has clarified the policy Chapter 6 trade-offs in the bailout of countries. Large bailout packages can substantially reduce or eliminate altogether the large economic costs of capital outflows, costs falling disproportionately on workers and the weakest sector of the economy. A reduction of the costs associated with default, however, does induce moral hazard distortions. Debtors may be less ready to insure against crises: they may hold an insufficient level of international reserves; and most importantly, they may postpone important reforms or avoid the implementation of good policies when these entail political costs. This is “debtor moral hazard”. As mentioned above, international investors may lend while paying insufficient attention to country risk, as they expect bailout packages from international institutions to help rescue the country. This is “creditor moral hazard”. As a result of past FDI, EU-based corporations own a large number of financial intermediaries and firms in acceding countries. With financial deepening in these countries, households in other EU countries may be expected to include significant holdings of acceding countries’ liabilities in their portfolios. In the event of a default crisis, this implies that some non-trivial part of its cost will fall on households and firms in other EU countries. Political-economy considerations will strongly influence the balance between “bailouts” and “bailins” (that is, how much international investors will lose) in the event of a crisis within the EU. In the 2003 report (Chapter 4) we presented evidence of the fiscal costs of debt crises in OECD economies. Past experience shows that these costs can be substantial: the risks for financial stability and the presence of euro-area-wide spillovers are likely to play a key role in the decision process. 5. Are acceding countries different from other emerging markets? Recent studies on currency and financial crises pointed out a set of crucial institutional and economic features that characterise emerging markets. A consensus list (as in Mishkin 2003) includes 1. 2. 3. 4. 5. weak fiscal institutions weak financial institutions low credibility of monetary institutions dollarisation/euroisation of liabilities overall, greater exposure to a sudden end of capital inflows and to liquidity crises. To what extent are acceding countries different from typical emerging markets? There are at least two notable differences: the ultimate goal of full integration into the EU as well as into the euro area and participation in a multilateral exchange rate agreement/international monetary system. The ultimate goal of full integration into the EU and into the euro area First, the stabilisation efforts of acceding countries are clearly driven and motivated by the ultimate goal of full integration into both the EU and the euro area. Not only does this endpoint constraint create a strong incentive for governments to pursue reforms – in terms of expected benefits of closer ties to Europe. It also provides a clearly defined agenda of institutional and policy reforms, influencing private sector expectations. In the accession process, fiscal, monetary and financial matters are subject to multilateral surveillance. Integration into the European financial markets could, in principle, reduce the exposure of these countries to liquidity crises. A similarly clear end-point constraint cannot be found – at least in an Box 6.2 Financial fragility and the sustainability of hard pegs equally strong form – in the experience of other emerging Gale and Vives (2002) provide a formal analysis of the costs and benefits of adopting hard currency boards (or unilateral euroisation) from a financial stability markets. perspective. Three features characterise countries that stand to gain from a hard currency board and therefore from giving up their monetary policy altogether: these are countries that have (a) a weak institutional structure (including lacking truly independent central banks), (b) serious but not extreme moral hazard problems in the private sector, and (c) moderate liquidation costs of business projects. For these countries, the cost of operating without a lender of last resort (LOLR) is smaller than the benefits from the commitment to fight inflation. Note that low moral hazard implies that risk-taking is not that excessive (the number of firms that would default in the event of macro or sectoral shocks is therefore smaller than otherwise). In the case of default, economic and social costs are contained. A national monetary policy is better for countries with a stronger institutional structure. In their paper, Gale and Vives present an analysis of a sample of countries including Turkey. The authors conclude that Turkey is a candidate for euroisation despite important liquidation costs. Note that concerns about the stability of the banking system are a reason to diversify the choice of roads to EMU across acceding countries. 129 Participation in a multilateral exchange rate arrangement Second, governments of the acceding countries expect to play some role in the decisionmaking process of the ECB and other EU institutions. The extent to which this translates into an effective influence is uncertain. EEAG Report Chapter 6 But no other emerging market in Latin America and Asia has any formal link with, say, the US Federal Reserve or the Bank of Japan. desirable in terms of “constructive ambiguity”. EU enlargement strengthens the case for a reform of the EU financial architecture. Specifically, participating in ERM II is different from unilaterally pegging a currency to the euro. Multilateral surveillance and integration into the European institutions strengthen the credibility of domestic monetary authorities. There are explicit mechanisms regulating liquidity provision in case of need. Yet, it is clear that liquidity support from the ECB will not be boundless but will be subordinated to maintaining price stability in the euro area. An often heard argument is that financial crises in any of the acceding countries are not a concern because each of them is economically very small relative to the EU. Thus, the argument goes, the risk of EU contagion is limited, and the EU can easily “afford” the costs of a regional crisis. This argument is not convincing. Even if the effects of a crisis in the richer regions of the EU can be contained, there could be “horizontal” contagion among new members. Acceding countries may be small in terms of GDP but not in terms of population. Financial contagion can create widespread harm, generating political sentiments against European integration. Second, EU countries may have different and possibly conflicting views about the appropriate EU policy in the event of a crisis. This may make the EU response slow, uncoordinated and eventually ineffective (even if large), with adverse effects on the magnitude of the crisis. Eventually, the costs of a crisis for EU firms and institutions may be significant, and their distribution across EU member states is bound to create political conflicts. The two differences above may mitigate the credibility problem stemming from weak fiscal and financial institutions as well as the credibility problem of monetary authorities (points 1, 2 and 3 in the list above). These differences however do not shield acceding countries from financial turmoil It is, however, unclear how institutions and markets in acceding countries will deal with financial stress if and when it comes. Possible shocks can take the form of higher interest rates with demand growth expansion in Europe, reversal of capital flows driven by domestic or foreign events and strong fluctuations in the terms of trade and commodity prices – think of the implications of strong dollar depreciation. Some imbalances and shocks are likely to originate domestically, during the process of convergence, as relative price and structural adjustments may produce changes in the production structure. As argued above, EU accession is not likely to mitigate the problems raised by currency mismatch in foreign liabilities. If anything, one may expect capital inflows to intensify after EU accession that provides these countries with a “seal of approval.” Note that in the political debate, the issuance of euro-denominated liabilities could even be welcomed as a positive step towards full integration into the euro area. Stronger institutional ties with Europe could reinforce expectations of bailouts in the event of a crisis, leading to excessive risk-taking by both local and institutional investors. In our 2003 report (Chapter 4), we stressed that the presently ill-defined procedures for dealing with financial crises within the EU create considerable uncertainty about policy responses to a crisis, well beyond what could be EEAG Report 130 6. Preparing for the euro Acceding countries are expected to spend at least two years in ERM II prior to entering EMU. In light of the considerations above, two or more years in ERM II after accession could expose these countries to major currency and financial instabilities. Acceding countries’ intended strategies In view of the risks associated with regimes of limited exchange rate flexibility, all acceding countries have stated their intention to participate in ERM II for as short a time as possible, that is, no more than the two-year requirement, before entering EMU. They see ERM II as a “waiting room”, with no recognisable merit or contribution to the convergence process. Strategies, however, differ across countries as regards the timing of EMU entry. One group of countries aims at joining ERM II as early as possible after accession. These are the countries that already have hard pegs or currency boards, including Cyprus, Estonia, Latvia and Lithuania. Currency boards and Chapter 6 hard pegs to the euro as the reference currency have been declared compatible with participation in ERM II and therefore qualify a country for EMU membership (Governing Council of the ECB, April 13, 2000). Thus, these countries will not change their exchange rate regimes between EU accession and the adoption of the euro – some of them however have to revise the currency basket to which they peg, as to make the euro the only reference currency. For the group of countries that is instead pursuing some form of inflation targeting with flexible exchange rates, participation in ERM II is a clear change to an intermediate regime with only limited exchange rate flexibility before adopting the euro. They will therefore need to undertake a double regime switch, from the current regime to ERM II, and then from this to EMU. In some cases (the Czech Republic, the Slovak Republic and Poland) policymakers have expressed a preference for delaying ERM II participation for some time. This will give them time to achieve some progress in financial development, and, most importantly, to put their fiscal house in order. A relatively slow pace on the road to the euro could be appropriate for countries with mild fiscal problems – to the extent that mild fiscal problems are not priced in too harshly by international investors (i.e., to the extent that country and currency risk premia are not too high in international markets). However, countries with relatively bad fiscal fundamentals may also have a strong incentive to target EMU entry as early as possible, to achieve fast interest rate convergence with the euro area and in this way reduce the government’s interest bill. These countries are primarily worried about being exposed to fluctuations in risk and currency premia. This seems to be the reasoning underlying the strategy of Hungary, a country that is targeting early ERM participation despite apparent fiscal imbalances. The ECB view In accordance with the spirit and the letter of the Treaty of Maastricht, the ECB sees ERM II as a means to achieve nominal convergence and macroeconomic stability, and ultimately to foster real convergence and growth. More precisely, the ECB sees ERM II as a catalyst, enhancing the discipline of stabilisation policies and domestic policy institutions (ECB 2003). Consistent with this view, there is no reason to limit participation in ERM II to two years 131 only. The risks that we have discussed in this chapter are not inherent in the system: rather, they are inherent in premature participation in ERM II. The modalities to join should be decided on a case-bycase basis, looking at the progress of a country in implementing structural reforms, achieving policy credibility, and implementing stabilisation policies. Moreover, the ECB points out that real convergence may be associated with changes in the equilibrium real exchange rate, which are easier (less costly) to achieve via nominal realignment than domestic price adjustment. Once in ERM II or EMU, the adjustment will necessarily fall on prices. Joining ERM II and EMU at a later stage, after a country will have sufficiently advanced in the process of real convergence, may help reduce macroeconomic costs. The rationale of these different views of nominal convergence has been the focus of an intense debate. Specifically, there are strong concerns about three dimensions of nominal convergence: the ERM as a regime of intermediate exchange rate flexibility, the consistency of inflation and interest rate stability, and the extent of fiscal flexibility. Exchange rate stability As regards “exchange rate stability” as a criterion to qualify for EMU membership, an important issue is whether large exchange rate fluctuations within the official bilateral 15 percent band around central parity would be considered an indicator of “tension” in the exchange market, disqualifying a country from EMU participation. Will the criterion be applied with reference to a much narrower band, say 2.25 percent, the size of the band in the pre-1993 ERM? According to the ECB (2003), “the assessment of exchange rate stability against the euro will focus on the exchange rate being close to the central rate”. This issue may be a minor concern for the set of countries adopting currency boards and hard pegs. It is, however, crucial for countries currently using some form of inflation targeting and therefore in need of undertaking a regime switch prior to entering the euro area. In light of the recent financial history summarized in this chapter, the ERM with narrow bands is the kind of intermediate fixed exchange rate regime that invites speculation and makes countries vulnerable to severe liquidity shocks. If the convergence criterion requires countries de facto adopting narrow EEAG Report Chapter 6 bands, acceding countries will be forced to take unnecessary and useless risks. supporters of euroisation see the exchange rate as a potentially destabilising price (in the event of a crisis). Several documents of EU institutions and the Eurosystem seem to define “exchange rate stability” as an asymmetric criterion, that is, compatible with appreciation but not with depreciation. But acceding countries are recommended to set their initial parity in ERM II according to their best guess of the currency’s fundamental value, based on a broad range of indicators, including market prices, rather than “playing games” with an eye on the final euro conversion rate. It is apparent that such a recommendation is not consistent with an asymmetric definition of exchange rate stability. Clearly, such a definition provides a strong incentive for acceding countries to choose a relatively weak central parity at the beginning of their participation in ERM II and let their currency appreciate over time. It is not surprising to see acceding countries’ governments strongly arguing that their currencies are overvalued (while perhaps intervening heavily to prevent appreciation). A crucial dimension of this trade-off is the extent to which the elimination of exchange rate risk raises the default risk in the economy. With weak financial institutions and markets, immediate euroisation will possibly exacerbate moral hazard problems leading to excessive risk taking. An ill-defined financial architecture for the euro area as a whole may magnify the the distortion, offsetting the benefits of sheltering balance-sheets from valuation shocks due to exchange rate movements. We should note here that uncertainty about the final euro conversion rate could actually damage the country as well as the stability of ERM II by creating a coordination problem among market participants: with which final conversion rate would markets coordinate their expectations?4 The benefits from a clear endpoint exit from exchange-rate based stabilisation would in part be eroded. The (unfeasible) option of immediate euroisation Immediate euroisation, even in the form of unilateral adoption of the euro, would eliminate exchange rate risk and solve the problems raised by currency mismatches in the country balance sheets when external debt is denominated in a foreign currency. It would therefore close an important channel through which self-fulfilling prophecies in the exchange market and exchange rate crises can have devastating effects on the economy. Among the policy trade-offs of immediate euroisation, an important one concerns relative prices. An excessively appreciated (or depreciated) initial conversion rate between the domestic currency and the euro could create large and protracted real costs in terms of employment, investment and growth. But 4 ERM realignments raise a number of well-known policy issues as regards their timing, size and guidance of market expectations (see Buiter, Corsetti and Pesenti 1998 for a discussion). EEAG Report 132 Moreover, in the political economy of EU accession, participating in EMU is a well-understood ultimate goal that can motivate reforms and good policy making in the third phase of the accession process. Immediate euroisation will substantially reduce the leverage of current EU members on acceding countries, as well as of domestic governments on domestic and international interest groups. As discussed in Chapter 1 of this report (Appendix 4 on the Past and Future of the Stability and Growth Pact), the goal of entering EMU can motivate large fiscal consolidation efforts, but common fiscal rules become much less binding once a country is in EMU. The European Council in Nice, however, excluded euroisation from the set of relevant policy options open to acceding countries on the ground that it would be inconsistent with the view underlying EMU as the endpoint of a convergence process, adopted by the Treaty of Maastricht (Council of the European Union Press Release No. 13055/00; see also European Central Bank 2003). Euroisation is seen as a way to circumvent the convergence process. Currency boards Would an early adoption of a currency board provide a good substitute for early euroisation to address the issue raised by the currency denomination of foreign debt and the escape clauses implicit in intermediate regimes of fixed exchange rates? There are strong reasons to be sceptical. First, markets may still attach some positive probability to devaluation. Second, (as in the case of unilateral euroisation) the ECB will not be required to act as de facto lender of last resort (although it may choose to do so if there is no danger for its price stability objective). Third, we have seen that many countries are suffering a Chapter 6 deterioration of their fiscal state. Large deficits are not compatible with adopting a currency board. As experience shows, a currency board is not per se effective in forcing convergence of the country risk premium: interest rates may not fall at all if markets are not absolutely convinced about the sustainability of the fiscal and financial systems. These are among the reasons why a successful currency board requires a country to meet strict fiscal and macroeconomic conditions. If readiness for the euro is to be judged in terms of a country’s ability to sustain a currency board, the accession to the euro will be a longer and more risky process than most acceding countries would hope for. Inflation versus exchange rate stability The second dimension of the convergence process under scrutiny concerns inflation. Given that acceding countries are growing at fast rates, and the price level tends to increase with income, is there a conflict between exchange rate stability and price stability? Many observers trace a possible conflict between exchange rate and price stability to the BalassaSamuelson theory. The well-known argument is as follows. Fixing the exchange rate pins down the price of tradables in domestic currency. As gains in productivity in the tradable sector cannot translate into lower prices, they translate into higher wages that, with sufficient labour mobility, will spread across industries in the non-tradable sector. But these sectors experience much lower productivity growth: higher wages can only be paid if the price of nontradable output goes up. A high rate of price increase for non-tradable goods (which is an equilibrium relative price adjustment) may raise observed CPI inflation above the convergence criterion. Suppose the Balassa-Samuelson effect was indeed the main determinant of inflation and real exchange rate appreciation in acceding countries. If monetary policy targets some low inflation rate, such policy would result in some moderate appreciation of the exchange rate. According to the Balassa-Samuelson theory, an appreciating exchange rate would translate into a fall in the domestic price of tradables relative to non-tradables, a fall that is completely offset by gains in productivity. Thus, choosing an inflation target would not violate the exchange rate stability criterion (as the exchange rate would appreciate) and at the same time would produce a fall in the price of tradables in domestic currency (not to be confused with deflation). 133 Conversely, if a country pegs the exchange rate, there would be some inflation differentials during the income convergence process. Available empirical studies produce a wide array of estimates of the size of inflation differentials attributable to BalassaSamuelson effects. While most studies predict small differentials, unlikely to cause violation of the Maastricht inflation criterion, there are also much higher estimates. Critics of the convergence criteria point out that, in the presence of nominal rigidities in the economy, a binding inflation criterion would only produce unnecessary harm, as it would confuse equilibrium adjustment of relative prices with a general increase in the price level.5 Thus, according to these critics, the inflation criteria should be made more flexible for countries that choose hard pegs or currency boards. To sum up: during real convergence, sticking to a low inflation target would imply a moderate rate of exchange rate appreciation (which has been declared consistent with the exchange rate convergence criterion); adopting a hard peg would imply a rate of inflation which could violate the inflation convergence criterion. In principle, this criterion should be relaxed. Are there specific reasons to prefer one regime over the other? Inflation differentials depend on much more than Balassa-Samuelson effects Even in the case that, at some stage of the convergence process, inflation differentials could be entirely attributed to desirable relative price adjustment, relaxing the inflation convergence criterion is risky. This is because high inflation rates could feed agents’ expectations of further price dynamics. The potential problems raised by such expectations include overvaluation of the real exchange rate and short-term excessive demand expansion. This argument stresses that in reality inflation and real appreciation during the convergence process reflect much more than the Balassa-Samuelson effect. There are also (a) changes in the terms of trade; (b) changes in the size of deviations from the law of one price for tradable goods; and (c) measurement errors that may be sizeable in acceding countries due to rapidly changing baskets of consumption goods and the structure of production. In 5 The problem could be mitigated by calculating the reference value for inflation using the euro area inflation rate, rather than an average for three countries with the lowest inflation in the EU. EEAG Report Chapter 6 menting many reforms that could require temporary government spending. Given the political and institutional crisis concerning the SGP, critical views are likely to gain strength over time. Figure 6.3 addition there could be (d) inflationary effects of domestic over-heating caused by capital inflows and private sector expectations. There is insufficient understanding of the relative importance of these elements in practice. Figure 6.3 plots the ratio between the producer price index (PPP) and the consumer price index (CPI) in the different accession countries. It is reasonable to expect that the PPI includes a larger share of tradable goods than the CPI, as services have a larger share in the latter. In that case, a strong Balassa-Samuelson effect would imply a fall in the PPP relative to the CPI, since a relative increase in the price of non-tradables would raisethe CPI more than the PPI. The figure shows that a pattern consistent with the BalassaSamuelson hypothesis can be detected for a few, but not all acceding countries. There are large differences in real exchange rate patterns. Overall, these considerations suggest caution in relaxing the inflation criterion as a precondition to participation in EMU. Overall, a regime of (some degree of) exchange rate flexibility coupled with low inflation dynamics seem to provide a safer path to joining EMU. The need for fiscal discipline and the Stability and Growth Pact As regards the final concern with convergence criteria, some observers claim that the Stability and Growth Pact (SGP) is likely to become a straitjacket for fast growing countries at low levels of income, in need of building their own infrastructure and imple- EEAG Report 134 It is quite obvious that fast-growing countries are in need of large public investment, and that there are reasons why it is neither optimal not equitable to finance infrastructure with current taxes. By the same token, institutional and structural reforms may create temporary spending or revenue shortfalls, which would be optimally financed by borrowing. Yet one needs to be aware of the fact that local governments have a strong incentive to use fiscal flexibility for other purposes than the accommodion of a high rate of public investment or the facilitation of reforms (see the box on the SGP in Chapter 1 of this report). At times when governments are making an effort to consolidate their budgets and find viable fiscal paths, a relaxation of the fiscal rules may compromise the consolidation process, with negative effects on expectations affecting interest rates on the public debt. Our proposal for an improved Stability and Growth Pact (summarised in Chapter 1 of this report and extensively discussed in Chapter 2 of the 2003 EEAG report) can, however, suggest ways to grant some degree of flexibility in the design of fiscal policies. The enlarged euro area Many of the problems discussed in this chapter will still persist at the end of the third phase of EU accession, when the euro area will consist of at least 22 countries. Enlargement obviously exacerbates issues regarding efficient monetary policy decisionmaking in the euro area: a very large decision-making body cannot work well. These issues will call for a much needed structural reform, which we do not see as a major issue, however. After all, monetary policy in the euro area is not a weighted average of national policies. Yet, enlargement raises new policy issues. In an enlarged euro area, the acceding countries will be small in terms of economic weight but not in terms of population. Chapter 6 A small economic size means that acceding countries will not have much weight in the design of stabilisation policies. Even if inflation were tending to be higher in the new countries than in the current euro area, their influence on interest rate setting by the ECB would be negligible. First, their weight in the euro area Harmonized Index of Consumer Prices is very small. Second, a significant part of any inflation differentials is likely to be attributed to desirable relative price adjustments. A large population share in the EU, however, implies that the importance of acceding countries may become much larger in the event of financial turmoil and crises. Markets may expect political considerations to guide EU crisis management and resolution. Even if the ECB were to be able to pursue a consistent monetary policy through periods of turmoil, crises would be formidable challenges to national fiscal authorities and financial supervisors/regulators. As argued in chapter 4 of last year’s report (EEAG 2003), defining procedures and intervention policies in the event of crisis at the euro area level may be extremely helpful in this respect. 2004, and Latvia and Slovenia in 2005. In this group of countries, Hungary has the most fragile fiscal situation. The other countries have chosen to delay their entry into ERM II, perhaps waiting to see whether the interpretation of the convergence criteria will be adapted in light of their arguments against the merits of ERM II. Countries that are already able to sustain hard pegs should be helped to achieve a smooth and fast transition to the euro. In this set of countries, mainly small ones, priority should be given to institutional reforms and to building a policy framework consistent with participation in the euro area without suffering from major macroeconomic imbalance. Delaying participation in ERM II is a realistic option for countries that are currently unable to sustain hard pegs and have large domestic imbalances. The magnitude of domestic imbalances varies considerably across countries, so that ERM entry may be desirable at different times. Yet in all cases, the policy priority is achieving a sustainable fiscal situation and stabilising inflation at the correct relative prices, a task that requires both institutional and policy reforms. 7. Summary Overall, there is no single strategy that could be recommended to all acceding countries as regards macroeconomic stabilisation on the road to the euro. Arguments in favour of adopting the euro as early as possible include smaller financial risk due to the elimination of currency mismatch in the balance sheet of banks and firms (which implies the risk of a self-fulfilling run on the country debt); interest rate convergence (with the associated gains in terms of the interest bill for the government as well as investment financing by firms); and overall gains in monetary credibility. Arguments for a slower pace toward the euro stress the need to remove financial distortions creating moral hazard and therefore undermining the stability of the domestic financial sector and raising the country’s default risk; the advantage of relative price adjustments without the need of costly nominal wage and price adjustments; and the need to make fiscal and financial policy sustainable and compatible with a fixed exchange rate before participation in the EMU. At the end of 2003, some countries reiterated their willingness to enter ERM II on or shortly after accession: Hungary, Estonia, Cyprus, Lithuania in 135 For both groups of countries, the convergence criteria in terms of inflation, interest rates, debt and deficit provide desirable targets to guide policy and should not be relaxed. Though they are not first-best targets, these convergence criteria should be judged relative to existing distortions that could derail the stabilisation efforts. The evidence reviewed in the previous chapter shows that on average acceding countries are doing well as regards the two criteria of inflation and longterm interest rates. Once in the EU, it is even possible that the three countries with the lowest inflation rates will include acceding countries. The main issue is fiscal convergence (which of course may undermine the sustainability of the inflation and interest rate performance). As regards exchange rate stability, ERM II allows for large fluctuation bands around exchange rate parity. Once in ERM, a country should be able to use the exchange rate flexibility implied by such an arrangement, in the sense that exchange rate stability should not be mechanically assessed with reference to much narrower bands. Fluctuations in the exchange rate in response to domestic and foreign EEAG Report Chapter 6 shocks are not necessarily indicators of tension in the exchange market but can be part of an efficient adjustment process. If the dollar continues to depreciate, it may be reasonable to expect exchange rate fluctuations within ERM II. Declaring that acceding countries will be accepted in the euro area only if they can peg to the euro within narrow bands may raise the possibility of speculative attacks driven by self-fulfilling prophecies. During the transition to the euro, strict domestic stabilisation with some exchange rate flexibility is better than exchangerate-based stabilisation with very limited flexibility. In practice, however, exchange rate flexibility will not be enough to shelter a country from financial turmoil. The risk of crisis is somewhat reduced, but not eliminated, by delaying participation in ERM II, or by making full use of the 15 percent bilateral bands once a country is part of the ERM system. “Convergence play” and currency mismatches can still characterise the transition to the euro. A dangerous possibility during the transition is that markets do not learn to appreciate countries’ specific features and assess country risk based on domestic policy and real fundamentals. It would be extremely frustrating if the policy effort of one country were to be discounted in episodes of turmoil, whereas markets extend to all acceding countries the adverse assessment of a subset of them. The 1992–93 experience of the ERM as a multilateral system clearly shows that financial contagion is possible even in advanced countries (see Buiter, Corsetti and Pesenti 1998). The weakness of the old ERM are present, in magnified form, in ERM II. The experience with the ERM also points out that there is no stability without a consistent currency and financial policy framework for Europe as a whole. References Begg, D., B. Eichengreen, L. Halpern, J. von Hagen and C. Wyplosz (2003), “Sustainable Regimes of Capital Movements in Accession Countries”, CEPR Policy Paper no. 10. Buiter, W., G. Corsetti. and P. Pesenti (1998), Financial Markets and European Monetary Cooperation, Cambridge University Press, Cambridge, UK. Buiter, W. and C. Grafe (2002), “Anchor, Float or Abandon Ship: Exchange Rate Regimes for the Accession Countries”, Banca Nazionale del Lavoro Quartely Review, no. 221, June. Calvo, G. (2003), “The Mirage of Exchange Rate Regimes for Emerging Market Countries”, NBER Working Paper no. 9808. Corsetti, G., P. Pesenti and N. Roubini (1999), “Paper Tigers: A Model of the Asian Crisis”, European Economic Review 43, 1211–36. Corsetti, G. and P. Pesenti (2002), “Endogenous Optimum Currency Area”, NBER Working Paper no. 8783. EEAG Report 136 European Central Bank (2002), Review of the International Role of the Euro, December, Frankfurt. European Central Bank (2003), “Policy Position of the Governing Council of the European Central Bank on Exchange Rate Issues Relating to the Acceding Countries”, http://www.ecb.int, (found 18 December 2003). European Commission (2003), “Key structural challenges in the acceding countries: the integration of the acceding countries into the Community’s economic policy co-ordination process”, Occasional Paper, Directorate-General for Economic and Financial Affairs. European Economic Advisory Group at CESifo (2003), Report on the European Economy 2003, CESifo, Munich. Gale, D. and X. Vives (2002), “Dollarization, Bailouts, and the Stability of the Banking System”, Quarterly Journal of Economics 117, 467–502. Guimaraes, B. and S. Morris (2003), “Risk and Wealth in a Model of Self-fulfilling Currency Crises”, Cowles Foundation Discussion Paper no. 1433. IMF (2003), International Financial Statistics, Washington DC. Mishkin, F. (2003), Comments on Fraga, Goldfajn and Minella, “Inflation Targeting in Emerging Market Economy”, http://www.nber.org/books/macro18/mishkin8-14-03comment.pdf. OECD (1999), Bank Profitability: Financial Statements of Banks, database: www.sourceoecd.org. Tornell, A. and F. Westermann (2002), “Boom-Bust Cycles: Facts and Explanation”, IMF Staff Papers 49 (special issue). Tornell, A., F. Westermann and L. Martinez (2004), “Liberalization, Growth and Financial Crisis: Lessons from Mexico and the Developing World”, Brookings Papers on Economic Activity, forthcoming. Von Hagen, J. and J. Zhou (2002), “De Facto and Official Exchange Rate Regimes in Transition Economies”, Center for European Integration Studies Working Paper B 13.